The long-term goal of this project is to understand the combinatorial rules that govern the interactions between transcription factor binding sites (TFBS). Through these rules, combinations of TFBS specify an enormous diversity of gene expression patterns. Normal growth and development depends on the tight control of TFBS over levels of gene expression in both time and space, and aberrant regulation of gene expression underlies many genetic diseases. Although much progress has been made identifying TFBS, and the transcription factors (TFs) that bind to them, much less is known about how TFBS interact with each other to generate specific patterns of gene expression. In particular we lack an understanding of how specificity is achieved in large genomes where the cell must distinguish between true cis-regulatory elements and thousands of similar, but non-functional sequences. The proposal addresses this problem through experiments designed to unravel the mechanisms that govern cis-regulatory element specificity. Large libraries of simplified synthetic promoters will be constructed and assayed for expression, TF occupancy, and chromatin state. The data from these libraries will be analyzed with a thermodynamic model that describes the physical interactions between TFBS, TFs, RNA polymerase, and other elements that determine cis- regulatory activity. The models we produce from synthetic promoters will be explicitly tested on genomic promoters. The three aims of the proposal are organized around three different sources of information that determine specificity. In Aim1 we test the hypothesis that functional TFBS are specified by their interactions with other nearby TFBS. In Aim2 we explore the extent to which local sequence features besides TFBS help specify functional cis-regulatory elements. Finally, in Aim3 we will test the extent to which the chromosomal position and chromatin state of cis-regulatory elements influence the interactions between TFBS that govern their activity. The successful completion of the aims of this proposal will result in a quantitative and molecular understanding of the rules underlying combinatorial cis-regulation. Such an understanding is necessary to empower biomedical applications, such as stem cell engineering, that are based on manipulating gene expression patterns. The results produced from this proposal will also help guide the annotation of the large regions of non-coding DNA in the genome that specify gene expression patterns. Finally, a clear understanding of TFBS interactions will help the identification and interpretation of disease causing genetic variants that affect cis-regulation.